Project Summary. This proposal describes a five year career development plan for Soumya Raychaudhuri in statistical genetics. Dr. Raychaudhuri is a Rheumatology fellow at the Brigham and Women's Hospital (BWH). He will integrate his background in bioinformatics with the statistical genetics resources of the Broad Institute and the immunological and clinical strengths of BWH. Dr. Raychaudhuri will be mentored by Mark Daly, an associate and director for the Computational Biology lab of the Medical and Population Genetics Program at Broad;Dr. Daly is a recognized expert in the statistical genetics of auto-immune disease with a strong mentoring track record. Dr. Raychaudhuri will work closely with David Altshuler, Peter Gregerson, Dan Solomon, and Robert Plenge, and receive from them general career advice and specific scientific guidance on the completion of the proposed project. He will work with Lars Klareskog and Paul de Bakker in collaboration to help genotype and analyze cohorts of rheumatoid arthritis (RA) patients. The research program will emphasize the genetics of RA, with a goal of identifying and replicating new loci that confer increased disease risk. Detection of RA genes has been difficult since it is a polygenic disease that probably involves modest effects from many genes with complex interactions. Genetic studies have thus far identified PTPN22, TNFAIP3, and STAT4 as replicable loci outside the Major Histocompatibility Complex. We hypothesize that some susceptibility alleles are common variants. Therefore, to identify unrecognized loci, we focus our efforts on the combined analysis of three genome-wide association scans (GWAS) in RA. Each individual study may be too small to detect necessary effects, however. We further hypothesize that involved genes may be in common pathways or share biological processes. Computational analysis of functional genomics data (gene expression, protein interaction, scientific text) may be able to elucidate relationships. Testing functionally related gene variants in concert for disease association may increase the power of genetics approaches. We propose: (1) developing and applying new methods to combine small RA GWAS to increase power, and (2) developing and applying novel bioinformatics approaches to integrate functional genomics data into statistical genetic analysis. Relevance. The research proposed here aspires to find disease genes in RA using bioinformatics approaches with population genetics data. The identification of these genes has implications for (1) the rapid identification and diagnosis of RA, (2) recognizing critical pathways in the pathogenesis of the disease, and (3) defining future pharmaceutical targets.